HOW SAFE IS SAFE? 0</f\'?t^ac^ir^ r^^ '^^f ^" 2 4 !07P HOW SAFE IS SAFE? by D£ivid A. Lucht Deputy Administrator National Fire Prevention and Control ;\diT;ini strat.i on U.S. DeDartrr.ent of Coipjr'.erco A^<^^^-^ }/?^^,fy.^ /^^V,^^ ^^^. /^7 /^?K /r^ri HOW SAFE IS SAFE? . by David A. Lucht How safe are we? How safe do we vjant to be? Hov; do we measure safety? Hov/ do we decide? While these ire fundamental questions, they are asked all too infrequently in the fire safety field. Even v;hen they are asked, practically no one can ansv/er them. VJhen fire protection experts do step forr-zard to offer answers, their views are frequently received with question or doubt. Rexford Wilson, a leading fire protection engineer, recently put it this v.'ay: "We are perceived to be dealing in beads, trinkets and shells, whi'^^^ other professions are using mathe- matics, science and computers." V7hether or not we are consciously asking ourselves the important, basic safety questions, v/e are making safety decisions everyday. The decisions are made in many ways. Decisions are made by the national organizations which produce model codes and standards; decisions are made by the local officials who adopt those model codes and standards; decisions are made by the people v;ho design our buildings. • When m-jdel codes and standards are v/ritten by such organ- izations as the Building Officials and Code Administrators Inter- national (BOCA), the Southern Building Code Congress (SBCC) , the International Conference of Building Officials (ICBO) , or the National Fire Protection Association (NFPA) , they are v:ritten through the "consensus standards-making process." This means that people with varying viev7S and backgrounds come together to work out V7hat the substance and wording of the code or standard will be. Once agreeraent is reached, the text is submitted to a vote to formaliJie the consensus. The model code or model standard is then publislied for Federal, state or local consideration and adoption. By the time this process is completed, many decisions have been made; decisions as to hov7 safe V7e are and how safe we are going to be. Officials at the Federal, state and local level are responsible for the lega]. adoption of the model codes and standards Often these officials simply adopt by reference t-he consensus achieved by the model code organizations. But sometimes changes are made. The factors v;hich influence these changes are varied. They are often made in an environment or high emotion. Follov/ing a nursing hojp.e disaster in the state of Ohio, for example, fire experts testified about whait needed to be done to make nursing hom.es safer. SomiO s^^-id that plastic v/as tebaskets should be outlav/ed. Some said that carpeting should be outlav;ed. Some said that smoke detectors should be required because they respond quicker than sprinklers. SorAe said sprinklers should be required because they put the fire out. Some said that sprinklers should not be required because the v.'ater from the sprinkler system might make things worse rather than better. /ifter government officials hear testimony such as this, they make their decisions. And these decisions are the ones that determine just how safe we v;ill be. Engineers and architects also ma?:e decisions that e.-ffect hov; safe V7e v.'ill be. Their design decisions are often based upon the minimum requirem.ants of the codes in effect in their jurisdiction. Somietim.es their design decisions are m.ade v.'ith the benefit of consultation from a qualified fire protection engineer. Sometimes that fire protection engineer performs a systematic analysis of the building being designed to determine appropriate levels of safety. Sy;:tem.atic fire safety analysis such as this is rare, hov.'ever. If v/e are really going to reduce fire losses in the United States so that v/e are no long-r a leadincj nation in per capita dollar loss and death rates, v;e must get ourselves out of the vjorld of guessv.'ork and emotion. Wo must entor into an era of "mathematics, science and computers"; a logical fire safety design method. We have to begin making more objective decisions about how safe we are going to be. More objective decisionrnalcing based on hard data analysis and a systematic v;eighing of alternatives must take place in the v;riting of our model codes and standards, in the adoption of those codes and standards, and iTi the design of our buildings. LOGICAL FIRESAFETY DESIGN METHOD Some fire experts claim to have a logical fire safety design methodology v/hich enables them to determine hov; safe v;e are and to measure fire safety. The nuinber of those fire experts is small and sometim.es their analysis is viev;ed v;it.h doubt. Doubt, is sometimes expressed by local building offici^^ls or fire offic£ils vjho are not aware of the fire safety design method or are not con'v'inced of its validity. A logical fire safety design method has not bec-n thoroughly documented and judged by a legitimate nationwide professional orgdinization . The first systematic method for measuring fire safety was developed in 1902. In that year, A.F. Dean of the Western Board of Fire Underwriters published, "The Analytic System for the Measurement of Relative Fire Hazards." Dean's system assigned relative values to various fire protection characteristics of buildings. This system enabled a numerical evaluation of each building for fire insurance rating purposes. Using this scheme, the property insurance industry has determined fire insurance rates for different classes of buildings and uses with a variety of different fire hazards present. Using these rates, the industry has been / able to derive fire insurance premiums v;hich are adequate to cover operating expenses, loss payments, and profits through the years. Descendants of this analytical system are still being used today throughout the United States. The International Conference on Fire Safety in High Rise Buildings, convened in April 1971 by the General Services Administration (GSA) , Public Buildings Service, brought together 70 participants representing the many disciplines associated v/ith fire safety design and planning. The summary report, of that conference contains the follovzing significant statement: 1 "International Conference on Fire Safety in High Rise Buildings", April 12 - 16, 19 71, Air lie House, VJarrenton, VA. General Services Administration, Washington, D.C. , 1971. "What follov7S (in the tot .1 report) is the beginning of a systems concept jelled from the deliberations of the conference for addressing the firesafety element in high-rise buildings. The ultimate objective is a system v/hich can be integrated into a total systems design and management concept, considering functional goals and all required attributes of such buildings. This can be done by constructing a firesafe design, and management system, identifying all major elements in that system, presenting recomm.endations for satisfying elemental requireraents--including alternatives v;here they are known to exist--and by identifying a total system concept vvithin v;hic'; new solutions rriay be sought and gaps in knowledge filled." In 1972 the General Services Administration published its "system.s approach" in the "Interim Guide to Goal Oriented Systems Approach to Building Fire Safety' for use in determining fire safety criteria for governmental buildings. Here the quantification of fire safety depends on a series of probability curves which describe GSA's fire safety goals and loss experience, The GSA curves are subjectively developed. This is the only 2 "Interim Guide for Goal Oriented Systems ApproacTi to Building Fire Safety," General Services Administration, Washington, D.C April 19 72. complete analytical system for probabilistic evaluation of the anticipated success in building fire safety performance. The National Fire Protection Association created its Committee on Systems Concepts for Fire Protection in Structures, in June, 19 72, making it "responsible for developing systems concepts and criteria for fire protection in structures." Based on work done by the National Bureau of Standards and the General Services Administration, the coirumittee developed a "decision tree" which 3 was published in 1974. The NFPA and the GSA decision trees describe relationships between elements which make up a building's firesafetv system. Both show the hierarchy of conceivable factors and whether these act dependently or independently in the fire safety system. Each approach c:,n be quantified. The NFPA decision tree is the moi'e general form.. Today there are several variations of the systems approach in existence but few people know hov/ to use tiiem. What is v/orse, there is still debate among fire professionals as to which approach is better or in fact whotlier any of them are valid in the first place. P^s mentioned earlier, none of the firesafety system.s anaylsis methods have been thoroughly documented and evaluated by nationally recognized nrofessional associations. 3 "Decision Tree," Conmiittee on Systems Concepts for Fi.re Protection in Structures, copyrighted Noveml^er 19 74, National Fire Protection Association, Boston, Mass. WHERE DO VJE GO FROM HERE? If V7e are going to have an irr.pact on reducing fire lo:..;e3, a good place to start is v;ith an ob;;ective approach to questions such as how safe are v;e ani hov/ safe do v;e want to be? The National ire Prevention and Control Administration (NFPCA) is acting nov; to determine v;r.at VJe know and v^fhat we don't know in the area of systematic analysis of building fire safety. We vzant to identify v.'hf-t can be measured and what c^m't be measured. We want to determine tlie limits of the systems approach or the "decision tree." Hov7 can these approaches be used subjectively and ho.v can they be used objectively? NFPCA is v.'orking with the Society of Fire Protection Engineers (SFPE) to achieve this goal. A grant to the SFPE v;ill support a project to assemble the current body of knovzledge in fire protection engineeri': -- a systematic approach to identify and solve building fire safety problem.s. The result of this project will be a textbook describing the state of the art in building firesafety design. The text vjill include a description of the theory, approach and concept of a systems analysis procedure as well as fire protectior. engineering fundamentals. This project v;ill not include research to develop nev/ technology- Rather, it v/ill asserr.blr; the current state of the art into a cohei'ent and vinderstandable v/ritten form. The text v/ill not be a cookbook; it will address concepts, techniques, and problem-solving logic. Also, thie text will not desci"ibe how to design specific nechcinic^il components such as sprinkler systems. Rather, it will address the question of how one should go about deciding v/hether or not a sprinkler system should be part of the building firesafety system. The text will not cover hov: to design or specify ei fire V7all. It will describe a procedure to follow in deciding whether or not a fire v;all is needed as pai't of the building firesafety system. It is intended that this text will be a pi'actical, working tool for use in the real v;orld of solving today's firesafety problems. OTPIER FIRE .ADMINISTRATION EFFORTS The National Fire Administration has sevei'al other efforts underway to increase the use of system.s analysis in firesafety decisionmaking. T\-70 specific sections of the 1974 Fire Preventioi and Control Act address building fire safety. SectJ.on 12, 10 "Review of Codes," requires the NFPCA to reviev; fire prevention and building codes and "suggest improvements." It is felt that the National Fire Administration can make a signific£.nt contribution in this area by investing resources in the development of systems analysis and cost benefit analysis techniques v/hich v/ill help both those v;ho write the codes and those who adopt and enforce them mtake more objective decisions. Section 13 of the Fire Prevention and Control Act is titled, "Fire Safety Effectiveness Statements." This Section of the Act is aimed at "consumer decisioniaaking. " Under this provision, the National Fire Administration vrill develop guidelines i'or m.easuring and docum.enting the relative safety provided by various types of buildings. On a voluntary basis, building ov;ners and managers could apply these guide- lines to derive firesafety m.easures for the public or the "building consumer." For example, som.eday the traveler may look at various hotel listings in his travel guide and see hotels classified as A, B, or C with respect to fire safety. If the traveler desires, he or she may decide v/hich hotel to patronize based on the degree of firesafety provided. Obviously, a systematic and objective methodology must bo 11 developed to derive these kinds of measvires. Preliminary v7ork to develop analytical techniques for both "reviev; of codes" and implementation of the Fire Safety E:f f ectiveness Statements Program is nov; underway through a grant to the ") Fire Protection Engineering Department at the Univ-::rsity of Maryland. Other work is also in progress to assist the building designer in firesafety decisionmaking. The NFPCA i-ecently published "A Fire Protection Primer for Architects." This Primer is probably one of the first of its kind in theit it introduces the architect to firesafety design consider^ltions in a form which the professional architect can most easily relate to. Also, the Fire Adm.inistration ' s National Fire __, Academy has a contract with the National Loss Control Service Corporation to develop a 30-hour firesafety design course for architects. This will be the first in a series of courses designed specifically for the building designer. It v;ill introduce the student to some of the more fundamental concepts of building firesafety problems and their solutions. It is expected that this course v;i.l 1 be ready in about six months ; it v.'ill be disseiiiinatod in cooperation v/ith the American Institute of Arcliitects. Lerup, Lars; Cronrath, David; and Liv, John Koh Chifing, "Learn From Fire: A Fire Prohection Primer for Architects. University of California, College of PJnvironmental Design, Architecture Life Safety Group, Eorkelcy, CA . 1977. Prepared for the National Fire Prevention and Control Administration, VJasb-Jngton , D.C., C.rant number 75008. 12 The Fire Administration' s" National Fire Data Center is also making an important contribution. The collection and analysis of fire statistics are necessary in objective decisionmaking. The most sophisticated systems analysis techniques require the use of numerical values in making probabilistic determinations of events. Unfortunately, the United States has been sorely lacking in standarized, "processable" fire statistics. Three years ago, no two states collected fire data using compatible terminology and coding systems. The National Fire Data Center is v;orking vigorously to increase standardization across existing data systems and to help states start up nev? systems. The Data Center adopted NFPA Standard No. 901, "Uniform. Coding for Fire Protection," as a standard dictionary of terms and numerical codes. At this time, 19 states are in various stages of starting up or converting to standardized data system.s. The Firo Administration is providing grants, computer soft- v?are, training and materials to the various states to help them get into this National Fire Incident Reporting System. As this system grov;s and matures, our ability to conduct quantitative analyses v/ill grov/ substantially. This v/ill 5 '^iniform Coding 'for Firo Protection." National Fire Protection Association, Boston, Mass. NFPA No. 901-1976. 13 greatly reinforce and improve our ability to make objective decisions about hov; safe V7e are and hov; safe vre v/ant to be. As a result of the v.'ork done so far, the Fire Data Center is publishing "Fire in the United States" v/hich represents the most comprehensive and in-depth study of the nation's fire problem ever to be done. H0V7 SAFE IS SAF/^? How safe is safe? Safe is as safe as v;e V7ant it to be. K' make decisions in this regard every day in the model codes and standards committees, in city councils amd state legislatures, and in the design offices of professional architects. The pi-oblem is we do not knov; hov; safe our decisions are. If successful, the NFPCA/SFPE project will tell us how safe our decisions are; it v;ill tell us v;hat parts of our decisions must be based on judgement and v;hat parts of our decisions can and should be based on fact. Hopefully, this project v/ill also tell us where thf- voids are and v:here further v/ork needs to be done to take raore of the guessv.'ork out of our important decisions v.'hich ultimately determine the safety of people irorr. fire. 6 "Fire in the United States: Deaths, Injuries, Dollar Loss and Incidents at the National, State and Local Levels." National Fire Prevention and Control Administration, Washington, D.C., 1978. Ji^O i'S^ DATE DUE NETC Library 16825 South Seton Ave. Emmitsbufg, MD 21727 NETCLRC 049729 ■yyr..